CAJ | 학술논문

提出一种抗浪涌电压干扰的电力设备故障挖掘算法，针对故障数据集合中的数据敏感性下降问题，运用故障类间数据的相似性，计算故障之间的类间相似特征。将相似度较高的数据分配到同一个类簇，在运用遗传算法进行特征分类寻优，保证最优解的唯一性，排除故障数据受到干扰造成的检测弊端。实验证明，该方法可以很好的解决浪涌干扰下的电力设备故障挖掘效率，与引入前相比，挖掘的结果更加准确，鲁棒性较强。
제출일충항랑용전압간우적전력설비고장알굴산법，침대고장수거집합중적수거민감성하강문제，운용고장류간수거적상사성，계산고장지간적류간상사특정。장상사도교고적수거분배도동일개류족，재운용유전산법진행특정분류심우，보증최우해적유일성，배제고장수거수도간우조성적검측폐단。실험증명，해방법가이흔호적해결랑용간우하적전력설비고장알굴효솔，여인입전상비，알굴적결과경가준학，로봉성교강。
An anti surge voltage power equipment fault mining algorithm, for fault data in a data set in question, using the fault class similarity between data, calculation of fault between the similar characteristics between classes. Will higher similarity data assigned to the same cluster, the use of genetic algorithms for feature classification optimization, ensure the uniqueness of the optimal solution, troubleshooting data caused by interference test. Experiments show that the method can well solve the surge under the interference of electric power equipment fault mining efficiency, compared with before the introduction, the mining results more accurate, with a strong robustness.